Glaciers of Alaska and Northwestern Canada are losing mass at one of the highest rates of any mountain glacier system globally. High-precision measurements from the Gravity Recovery and Climate Experiment (GRACE) satellite mission have revealed changes in the local gravitational field along the Gulf of Alaska due to changes in these ice masses since 2003. Previous efforts have spatially resolved these mass changes to 100 x 100 km grid cells or mass concentrations (mascons) as part of a global GRACE solution. While mass change estimates at the scale of entire mountain ranges (i.e. several geographically-grouped mascons) show strong temporal correlation to surface mass balance and air temperature, and while ice loss magnitudes for all Gulf of Alaska glaciers agree closely with geodetic estimates from ICESat, most GRACE-derived glacier mass loss magnitudes do not match ground observations at the level of individual mascons. In this study, we examine several approaches for partitioning the most recent GRACE mascon solution for glacier mass change along the Gulf of Alaska to individual mascons. We derive sets of scaling coefficients for every mascon, representing the local averages of different topographic or climatological characteristics, which essentially serve as different measures of continentality. These characteristics include mean ice elevation and distance from the coast, derived from the Randolph Glacier Inventory, and mean monthly temperature and precipitation, derived from the gridded climate product PRISM. Each set of scaling coefficients (representing each continentality index) is evaluated by comparing our derived timeseries' of mass change to independent estimates from available ground and remote sensing datasets. We focus our preliminary validation on mascons within the Juneau Icefield area in Southeast Alaska, for which we have independent constraints on mass change from hydrological models and laser altimetry, and which acts as a test case for future work extending to the Gulf of Alaska. Ultimately, this study investigates our ability to extract mass balance estimates from GRACE at the level of individual mascons, in order to better understand the timing and magnitude of subannual and long-term changes of remote Gulf of Alaska glaciers.
